The manipulation, introduction and ecology of field layer communities in broadleaved woodlands

A thesis submitted in partial fulfilment of the requirements of the University of Wolverhampton for the degree of Doctor of PhilosophyThis study was concerned with developing and investigating techniques to enhance the herb communities of secondary woodlands, particularly recently developed urban woodlands, by manipulating existing field layers, enriching them through the introduction of woodland herbs, and by examining aspects of the ecology of selected species which could contribute to their successful establishment. While the concern was primarily with practical aspects of woodland groundflora establishment it was conducted within the wider perspective of woodland reconstruction in which the aim is the development of woody and herbaceous communities whose functional and floristic characteristics as closely as possible resemble, or have the potential to resemble, those of ancient seminatural woodlands. In a comparison of the functional and floristic characteristics of the field layers of ancient semi-natural woodlands and urban secondary woodlands, the field layer of even the oldest secondary woodland, a 100 year old plantation, resembled those of the ancient woodlands only with respect to competitive and potentially invasive species whose presence could prevent the establishment of a characteristic shade flora. A small experimental woodland allowed detailed study of the influence of four litter and mulch treatments on the establishment and regeneration of five woodland herbs over a three year period. The frequency of Lamiastrum galeobdolon in 144 0.51m2 subplots doubled in that time to 67% and there were 202 new independent plants. The deeper litter and mulch treatments promoted the establishment of new plants; distribution was also influenced by finer scale spatial variations related to the degree of shade and the background vegetation. Oxalis acetosella increased its frequency in the subplots from the original 33% to over 80% in three years; the density of new plants was highest in the deeper oak litter plots. There was a steady diminution in the numbers of the Hyacintho ides non-scripta bulbs originally planted, but they gave rise to a population of seedlings, some of which flowered in 1994. The two sown species, Silene dioica and Digitalis purpurea repectively achieved the highest and lowest frequencies. Both were less successful in the deeper litter treatments than the shallow litter or the control plots. The beneficial effects of the litter and mulch treatments on the introduced herbs continued after their influence on the background vegetation had diminished. Seed populations sown at different times, stratified or otherwise, were remarkably resilient to exceptionally harsh climatic conditions. Although no plants were recorded in the first year after sowing, many species subsequently flowered and set seed. Where dense swards of mainly grassy vegetation persisted on sites of fairly high fertility and where the canopy was insufficient for its effective suppression, the use of a single application of herbicide improved the establishment rate of introduced herbs. However, some of the introduced herbs were able to exploit very small gaps in competitive vegetation even where no herbicide had been used. The benefits of rotavation, either on its own or in conjunction with herbicide, were short lived. Mature amenity mulches, such as woodchip, provided a suitable seed bed for introduced herbs at two sites, while the absence of either mulch or a dense cover of background vegetation at a third comparable site, resulted in lower rates of establishment. Experimental study of the germination and growth of woodland herbs confirmed that new mulches on their own provided a very hostile environment for seedling establishment, but that this could be ameliorated by sowing seeds in a layer of compost placed at an appropriate depth in the mulch. Experimental investigation of three woodland species, Silene dioica, Circaea lutetiana and Viola riviniana indicated that they had the capacity to survive combined low irradiance and mineral nutrient deficiency and to respond simply to relief from one constraint, or synergistically to relief from both. A transitory growth check at eight weeks in Circaea lutetiana and Viola riviniana, possibly linked to the onset of mycorrhizal infection, could make seedlings of these species vulnerable in resource-poor situations